Lapita Pottery from the Small Islands of North-East Malakula, Vanuatu: a Brief Overview and Implications

11 Lapita pottery from the small islands of north-east Malakula, Vanuatu: A brief overview and implications Stuart Bedford Abstract A series of well-preserved Lapita sites was first identified on the small islands of Uripiv, Wala, Atchin and Vao, Malakula, in northern Vanuatu in 2001–2002. Further excavation on Vao and particularly Uripiv continued until 2011. The pottery shows the standard similarities with Lapita pottery generally but also demonstrates the development of very distinctive regional and even island-specific variation in form and motif design during the Lapita period. It suggests very rapid change in pottery form and decoration soon after initial colonisation of the archipelago; an aspect largely masked by the radiocarbon chronology. It also confirms that regional diversification was well underway during the Lapita period itself. This may relate both to the potential that these communities came from different origin points further west and that even during a single generation a range of factors may have encouraged localisation in a range of practices including pottery production. Introduction Archaeological research into Lapita in general and pottery in particular has come a very long way since the 1970s when the distinctive dentate-stamped sherds were generally accepted as belonging to a tradition that continued for 1000 years or more in some regions and was broadly divided into two distinct Western and Eastern styles (Green 1979:43–44). The definition of the varied styles comprising different vessel forms and associated decoration was primarily categorised using chronological, decorative and geographical criteria, with three styles ultimately being identified in the 1980s, Far Western, Western and Eastern (Anson 1986; Green 1979). However, with increasing numbers of sites being excavated and increasing numbers of reconstructable vessels being identified from different regions, it became more difficult to assign new forms and designs into the three-style scheme. A modification was the concept of Lapita Provinces, essentially subgroups designed to break down somewhat the broad hegemonic geographic styles. A Southern Province comprising New Caledonia was initially proposed (Kirch 1997:72–73) and ultimately another three Provinces, Far Western-Western, Western-Central and Eastern, were added to encompass the Lapita distribution (Sand 2001:68). However, the concept of Provinces has never really gained much traction as what specifically constituted a Province apart from 226 Debating Lapita geographical designations was never particularly clear. The more generally accepted framework is that of Summerhayes (2000, 2001), where Lapita is defined primarily in decorative, vessel form and chronological terms, Early, Middle and Late. While this latter scheme is proving to be the most practical, it is also becoming increasingly obvious that these terms have limited wider distributional application and that it is often difficult to specifically define these styles. In terms of Lapita overall, it is Early Lapita, largely restricted to the Bismarcks, that remains the most clearly identifiable, consisting of the widest range of vessel forms with the most concentrated and fine dentate-stamped motifs (Sand 2015; Summerhayes 2000). Rare examples of key definitive aspects of Early Lapita can also be found as far east as Vanuatu and New Caledonia (Bedford 2015:37, Figure 3) but not beyond. Middle Lapita can be contrasted with Early Lapita and there are a series of motifs that can be identified as widespread across the Lapita distribution. There is a reduction in the number of vessel forms and less concentrated design motifs, but there are also indications that increasing regional variation is beginning to develop. Across the distribution, Late Lapita displays significant regional variation with a loss of the rigidly structured design motifs and standard vessel forms (Bedford 2015; Burley 2007; David et al. 2011; Sand et al. 2011) and therefore requires archipelago- and even island- specific definition. There does remain, however, indications of some level of interconnectedness across the distribution in the form of the widespread appearance of distinctive decorative techniques, such as shell impression, and motifs, such as the simple zigzag created through a rocking motion of a straight stamp. More recently and following the reconstruction of Lapita vessel forms and design motifs from a selection of sites located in both Near and Remote Oceania, Sand has argued that there was a ‘clear need to focus some of our studies on the details of Lapita variation, in order to highlight the diversity of the early Oceanic tradition’ (2015:162–165). Sand laments the lack of attention to the reconstruction of vessel forms from most Lapita sites (2015:162), but this has always been hampered by both the fragmentary condition of the vast number of recovered sherds and the often very limited areas of sites that are excavated, greatly reducing the possibility that a series of key sherds from single vessels, essential for full reconstruction, will be recovered. The poor preservation of Lapita sites and the generally small nature of the sherds are what led to the initial focus on motif comparisons (Anson 1986; Green 1979:43), which has continued to be a major area of research (Chiu and Sand 2005; Noury 2005, this volume). However, as highlighted by Burley and LeBlanc, there are dangers in continuing to place too much emphasis on quantitative comparison of motif suites as attributes, as the complexity of dentate-stamped application or position on a vessel can be equally suggestive of different periods and regional connections or divergence (2015:181; see also Chiu this volume). This paper addresses Sand’s call for detail on Lapita variation by focusing on pottery recovered from the small islands of north-east Malakula, northern Vanuatu (Figure 11.1). It presents a range of primarily dentate-decorated sherds and a more limited number of associated vessel forms that demonstrate both regional variation and lingering connections. It suggests that rapid regional change in pottery form and decoration, which is masked by both broad categorisations of Lapita pottery and radiocarbon dates, could be occurring across much of the Lapita distribution. terra australis 52 11. Lapita pottery from the small islands of north-east Malakula, Vanuatu 227 167o E 169o E Torres Is Ureparapara Mota Lava Vanua Lava Mota Banks Is 0 50 100 kilometres Merig Gaua N Mere Lava 15o S Maewo Santo Ambae Malo Pentecost Vao Ambrym Atchin Malakula Paama Wala Lopevi Rano Epi 17 o S Uripiv Efate Erromango 19o S Aniwa Futuna Tanna Aneityum Figure 11.1. Vanuatu and north-east Malakula. Source: Stuart Bedford. terra australis 52 228 Debating Lapita Sites and chronology Excavations on the islands of north-east Malakula commenced in 2001 as part of a research and training program in collaboration with the Vanuatu Cultural Centre, its staff and their network of ‘fieldworkers’.1 The islands are small, all less than 2 km2, low-lying and surrounded by fringing reefs, apart from sheltered sandy beaches on the south-western coasts. The islands are also subject to regular uplift (Taylor et al. 1980) and ashfall from the nearby volcanically active island of Ambrym (Robin et al. 1993). All of these aspects helped narrow down potential areas for finding colonising sites and guided the survey methodology that was applied. There was no sign of any Lapita pottery or other associated artefacts on the surface, and so it was determined that test pitting across the uplifted back beaches was required if sites were to be found. Ultimately Lapita settlements were found on Uripiv (2001), Wala, Atchin and Vao (2002) (Bedford 2003:152– 155). All were located above and parallel to the foreshore in sheltered beach zones facing the large island of Malakula. The sites on Uripiv and Vao proved to the best preserved and additional, more extensive excavations were carried out on Vao in 2003–2004 and on Uripiv in 2009–2011 (Bedford 2006a, 2007; Bedford et al. 2011). The sites were all generally very well-sealed and preserved through a combination of tectonic uplift, accumulated debris from later habitation in the same area and overlying tephra-laden and cyclonic sand deposits (Bedford 2007). The primary objective of the fieldwork once the Lapita sites had been found on these islands was to establish the extent of the sites and identify any spatial and temporal variation during Lapita and later-period occupation. Excavations in various locations across Uripiv, in conjunction with the regular inspection of newly dug construction features over many years, have also provided stratigraphic and chronological detail for the wider island landscape. The general stratigraphy at all four Lapita sites is remarkably similar (Figure 11.2), as is the Post-Lapita expansion and later settlement across the islands. The Lapita settlements are established directly on the beach terraces and ultimately expand to cover 3000– 4000 m2 in area as defined by midden deposits with a dentate-stamped component. Villages remain in these coastally orientated positions for around 1000 years, slowly expanding over time. A distinctive change in the stratigraphy of the sites begins at c. 2300 cal. BP, when coral gravel sourced from the foreshore begins to be brought into the sites to act as a free-draining surface for household compounds. These surfaces were continually renewed and in many areas of the sites this layer is up to 50 cm thick. Over a c. 500-year period these Post-Lapita settlements continue to expand in the same coastal zone to over 10 000 m2 in area, as defined through the imported coral gravel layer. There is substantial disruption at c. 1800 cal. BP with the massive Ambrym caldera-forming eruption that showers the whole region with ash (MaCall et al. 1970; Robin et al. 1993). Whether such an event required evacuation or not on the islands of Malakula has yet to be determined, but the eruption proved to be a temporary hiatus and people were immediately presented with a radically transformed landscape.

Load more